BRYX

Traffic Control Device Detection

This model analyzes highway construction project imagery to automatically detect and identify traffic flow control devices.

40 credits (with a subscription)

80 credits (without a subscription)

Version 1.2

Free Trial available!

Supervised Machine Learning

This composite vision model combines two distinct types of machine learning: object detection to detect non-linear devices like drums, cones, etc.; and instance segmentation to detect linear devices such as multiple jersey barriers and fences.

Use Case

Quickly verify total traffic control device counts for billing purposes with this model’s summary report of total counts for traffic control devices.

Fast Analysis

The model is currently trained to detect:

Barricades (Type 3)
Concrete barriers
Cones
Construction safety fences
Drums
Jersey barriers
Longitudinal channelizing devices
Screens
Stationary crash cushions
Temporary traffic control signs

Required Inputs

PNG, JPG, or JPEG files

Note: This model is trained on oblique imagery, collected downward to an approximate 45 degree angle to the ground.

Expected Outputs

Annotated images
A summary report in XLSX format

The Traffic Control Device Detection model provides an easy-to-use yet powerful way to identify the traffic flow control devices used in your highway construction project. Designed for engineers and other professionals involved in highway projects, the model allows you to quickly identify any missing or mispositioned traffic control devices and quickly replace, secure, or repair them as needed. As traffic control devices are a key element of any highway construction project, this model provides you with a speedy way to check on the overall safety of your project.

Note: Current mAP (Mean Average Precision) score is 62%. Results will continue to be refined as the model is trained on more data.

Cookie Statement

We collect cookies to improve your experience with BRYX.